diy solar

diy solar

HP Blade Server Power supply to charge 48v

offgriddave

Solar Enthusiast
Joined
Feb 14, 2020
Messages
225
Hi I am interested in upgrading my generator power supply to charge my 48v lion battery packs

After some suggestions here, I found this HP 3000w 48v power supply:

HP ESP120 51.4V 57A MAX

$99-120 on ebay

Given this goes to 51.4V I might be ok

Has anyone delt with this power supply before? Is there a better solution or bang for the buck?

 

Attachments

  • s-l1600.jpg
    s-l1600.jpg
    200.6 KB · Views: 6
I've never tried it before, but wouldn't a lithium battery just saturate the power supply the same way shorting it would?

There has to be some kind of current limiting involved.
 
That's a good question. I have 10 batteries to hook up to. Each "15A" 48v nominal. 750w. If I charge one at 3000w that's 4C! Obviously don't do that. If I put 10 together, 7500w at 3000w charge rate, that's .4C. Is that better?

Other than that I can only figure out some kind of regulated power module

this one, but it only goes to 50V at 15A:

or this one, but it only goes to 5A:

If I gang them up would it work? But that pushes my budget $$$ into a real charger. I'm trying to save money.
 
Might work, depends on what the PSU does. Some go into shutdown (They think it's a short or overload) -> Not suitable for charging

Others do have current limiting and can be used as a charger. It only lacks any charging profile so you have to manually shutdown them once the batteries are ful.
But most heavy-duty powersupplies have a remote on/off. Some shunts/meters like the Victron BMV do have a relay contact, so you can use that to enable charging when < certain SOC, and stop once a predefined SOC is reached

Other interesting options (price and availability is different depending on the country) is eg the Eltek Flatpack2 units, or some DC rectifiers from old routers
I have for example a 3U serverrack unit, which contains 3 2800W rectifiers, can output up to 60V (adjustable) at 150A total

The rack https://static6.arrow.com/aropdfcon...c2fa108f0818aba92ffdd560/apcms28powerrack.pdf

Note: Pay attention to the available grid. Some powersupplies require 240V circuits to archieve their rated power and cannot provide full power with 120V feeds.
 
Thanks!

I have a 240V feed from my generator. The generator is rated at 3500w for propane. The manual recommends staying within 95% no above or below for maximum efficency, so this puts me near 3000w.

So I might need a battery shut off? Or it will overload? I should research Victron BMV?

Thanks hopefully I can get some more info on this power supply, perhaps somebody who is using it can share their experience.
 
Server power supplies are less than ideal as current limiting is not designed with battery charging in mind.
Redundant server power supplies often have balancing function so they can actually adjust the output voltage/current on demand but control can be bit tricky. https://serverfault.com/questions/6...-redundant-power-supplies-balance-consumption

Maybe better to look for telecom 48v "DC rectifiers" that are sometimes basically same as battery charger.
I have bunch of Ascom 1900W dc rectifiers and they offer adjustable output voltage and current limit and they limit the current according to load.

Some thinking and hacking is needed with these also but they might be easier option than server power supplies.
 
Excuse me for a minute, I am not an expert on power supplies vs rectifier. When I hear rectifier, I think a bunch of diodes for converting AC into DC. But you would need a way to smooth it out? And going from 120v DC (I guess there's probably some magic way to go from 120v -> 60v by folding the wave in the circuit, again I'm kind of a dummy in this area but just using my imagination for a moment) .... would still need a 60v to 48v voltage regulator?

Am I anywhere near the technical description of a "rectifier" ?

Also for clarification ... current limiting. That would be a circuit that slowly reduces the amps to match the target charge voltage? (Like 52v or 54v) and it would show up as the amps slowly going down as the battery got to 100%?


... I am really going out of my area of knowledge on this, curious.
 
I have a 12.4V 85A server power supply that I routinely use to "charge" a battery (technically it's attached to a battery to supplement the battery's output). It operates just fine down to about 11.9V outputting 85A. Below 11.9V it will cut-off and overload. It is attached to a battery system with other charging sources, and it happily sits there doing nothing when the battery is pushed as high as 16V.

I have seen many constant voltage power supplies like this one that behave in this fashion. They put out their rated output at their set votlage, and they can provide that current at the set voltage and some amount below that.
 
Excuse me for a minute, I am not an expert on power supplies vs rectifier. When I hear rectifier, I think a bunch of diodes for converting AC into DC. But you would need a way to smooth it out? And going from 120v DC (I guess there's probably some magic way to go from 120v -> 60v by folding the wave in the circuit, again I'm kind of a dummy in this area but just using my imagination for a moment) .... would still need a 60v to 48v voltage regulator?

Am I anywhere near the technical description of a "rectifier" ?

Also for clarification ... current limiting. That would be a circuit that slowly reduces the amps to match the target charge voltage? (Like 52v or 54v) and it would show up as the amps slowly going down as the battery got to 100%?


... I am really going out of my area of knowledge on this, curious.
Yeah, rectifier can mean bunch of diodes but in telecom "slang" it is full-blown power supply with bells and whistles.

Some of these have really nice featuers, Ascom ASM 1900W for example can handle full rated load with passive cooling (no fans)
 
A bunch of diodes is unregulated.

AC voltages is just a RMS value, not the voltage from 0 to the peak of the sinewave

If a transformer outputs 12V, that's the rated RMS Mvoltage. But since AC is a sinewave, the peak of the sine is a higher voltage, SQRT(2) times the nominal voltage. That means a 12V unregulated powersupply outputs peaks up to 16.92V


For lead-acid that doesn't really matter. In fact, a transformer + diodes is the most basic charger there is (And cheap). And usually a higher voltage (14V or so for a 12V battery) the peak voltage will be even higher
With a 48V LFP setup, you might even exceed the fets max rated voltage in such setup, destroying the BMS

But for LFP... that (near) 17V can do damage, as eg @Will Prowse experienced in a similar setup with a cheap leadacid charger
I
 
Back
Top